S956
Clinical - Non-malignant & functional radiotherapy
ESTRO 2026
intralesional steroids/5-FU) may further improve control and merit prospective testing. Keywords: keloid; treatment interval; local control
Digital Poster 4712
Excision-to-radiotherapy interval and keloid local control: single-centre real-world experience (2017– 2024) Shih-fan Lai Radiation oncology, National Taiwan university cancer center, Taipei, Taiwan. Oncology, National Taiwan university hospital, Taipei, Taiwan Purpose/Objective: Early postoperative radiotherapy (RT) is advocated to reduce keloid recurrence, yet the optimal timing window remains uncertain. We tested whether initiating RT >48 h after excision is non-inferior to ≤ 48 h for 12-month local control (LC12), using a prespecified − 10% absolute risk-difference margin (delayed − early); 6-month local control (LC6) among lesions with ≥ 6-month follow-up served as a secondary endpoint. Material/Methods: Single-centre retrospective cohort with lesion-level analysis and patient-level clustering. All cases received a uniform 6-MeV electron technique. Timing was defined by the excision-to-RT interval: ≤ 48 h vs >48 h. The LC12 cohort included lesions with ascertainable 12-month status (n=102; ≤ 48 h: n= 81, >48 h: n= 21); the LC6 cohort included all lesions with follow-up ≥ 6 months (n=128; ≤ 48 h, n= 102, >48 h,n= 26). Outcomes were modelled using binomial regression with cluster- robust standard errors; the primary adjusted model included site and BED ₂ (with age/sex when available). Exploratory non-inferiority was evaluated via cluster/bootstrap resampling (150 replicates). Results: In the LC12 cohort (n=102; ≤ 48 h :n=81, >48 h:n=21), crude local control was 88.9% vs 76.2% ( ≤ 48 h vs >48 h). After adjustment (site, BED ₂ ; cluster-robust SEs), the predicted LC12 was 87.9% for ≤ 48 h and 81.7% for >48 h (RD = − 6.2%, RR ≈ 0.93). The pre-specified non- inferiority criterion ( − 10% RD) was not met: the cluster/bootstrap one-sided 95% lower bound for RD was − 34.7%, with precision constrained by the small delayed group (n=21). Considering LC6 in all lesions with ≥ 6-month follow-up (n=128; ≤ 48 h:n= 102, >48 h: n= 26), the adjusted predicted LC6 was 97.1% ( ≤ 48 h) vs 100.0% (>48 h), yielding RD = +2.9%; the non- inferiority criterion was satisfied (one-sided 95% lower bound +0.9% > − 10%). Conclusion: Initiating adjuvant RT >48 h after excision failed non- inferiority for 12-month local control, whereas 6- month control met the prespecified margin— supporting limited, pragmatic flexibility beyond 48 h for early outcomes but reinforcing ≤ 48 h as the preferred operational target for durable control. Adjunct measures (silicone taping/pressure; selective
Digital Poster 4895 Mitral and Aortic Valve Dose–Volume Burden in STAR for refractory VT Jakub Cvek 1 , Otakar Jiravsky 2 , Jiri Plasek 3 , Lukas Knybel 1 1 Oncology, LF OU and FNO, Ostrava, Czech Republic. 2 Cardiology, Podlesi Hospital, Trinec, Czech Republic. 3 Cardiology, LF OU and FNO, Ostrava, Czech Republic Purpose/Objective: Stereotactic arrhythmia radioablation (STAR) is increasingly used to treat refractory ventricular arrhythmias. To enhance safety, planning is moving toward explicit protection of intracardiac sub- structures. However, available guidance (e.g., AAPM, STOPSTORM, QUANTEC, etc.) does not specify dose– volume limits for cardiac valves. In addition to evaluating potentially radiation-related valve side effects, we quantified single-institution exposure to the aortic and mitral valves. Material/Methods: We retrospectively analyzed 45 consecutive STAR plans from a single institution. From DVH exports we recorded, for both aortic and mitral valves: Dmean, Dmax, D_0.035cc, D_0.5cc, and D_1cc. Results are reported as median [IQR]. Doses for valves were compared using paired Wilcoxon tests. We also report proportions exceeding pragmatic thresholds. Transthoracic echocardiography was used to compare valve function before and 1 year after STAR. The associations between toxicity grade and doses were evaluated using the nonparametric Kruskal–Wallis test. Results: Six progressions (13 %) of known mitral regurgitation before STAR, and no aortic valve side effects were observed. Among 6 cases, progression was grade 1 in 4 and grade 2 in 2 patients. Kruskal–Wallis testing showed no significant association between grade of side effect and mean dose (p=0.088), minimum dose (p=0.306), maximum dose (p>0.45), D0.035cc (p=0.681), D0.5cc (p=0.458), or D1cc (p=0.395). For the aortic valve, median Dmean, Dmax, D_0.035cc, D_0.5cc, and D_1cc were 0.96 Gy (0.41–2.28), 5.63 Gy (2.75–10.39), 5.12 Gy (2.69–10.10), 4.41 Gy (2.15–8.53), and 4.10 Gy (1.84–7.87), respectively. The corresponding values for the mitral valve were higher across all metrics—Dmean 1.74 Gy (0.78–3.58), Dmax 18.04 Gy (5.65–27.03), D_0.035cc 16.41 Gy (5.43– 26.16), D_0.5cc 12.38 Gy (3.56–20.99), and D_1cc 9.96 Gy (3.01–19.51). In terms of pragmatic thresholds, 42% of plans reached D_0.035cc ≥ 20 Gy to the mitral valve versus 18% to the aortic valve; for D_1cc ≥ 10 Gy, the
Made with FlippingBook - Share PDF online